IDENTIFIKASI COVID-19 BERDASARKAN X-RAY THORAX DENGAN METODE GLCM DAN PROBABILISTIC NEURAL NETWORK

Ramadhanty, Viony Karina (2022) IDENTIFIKASI COVID-19 BERDASARKAN X-RAY THORAX DENGAN METODE GLCM DAN PROBABILISTIC NEURAL NETWORK. Other thesis, UPN "Veteran" Yogyakarta.

Preview	Text daftar isi.pdf Download (101kB) | Preview
Preview	Text pengesahan.pdf Download (841kB) | Preview
Preview	Text Abstrak.pdf Download (120kB) | Preview
Preview	Text Cover.pdf Download (1MB) | Preview
Preview	Text Daftar Pustaka(2).pdf Download (111kB) | Preview
	Text LAPORAN-123160134-VIONYKARINARAMADHANTY.pdf Restricted to Repository staff only Download (5MB)

Abstract

ABSTRAK
Pandemi yang disebabkan oleh novel coronavirus telah mengubah perilaku setiap
individu diseluruh dunia. Mulai dari pembatasan sosial, bekerja dari rumah, memakai
masker saat berpergian serta membawa hand sanitizer. Virus corona ini menyebabkan
penyakit dengan gejala yang mirip dengan kasus pneumonia. Kemunculannya sebagai
penyakit baru serta kemiripan dengan pneumonia menyebabkan petugas medis kewalahan
dalam mengatasi pandemi ini. Secara umum pengujian dilakukan dengan teknik Real-Time
Reverse Transcription-Polymerase Chain Reaction (RT-PCR), namun identifikasi citra
radiologi juga dapat menjadi alat pertimbangan untuk pengujian Covid-19.
Penelitian ini menggunakan citra X-Ray Thorax untuk mengidentifikasi infeksi Covid-
19. Citra akan melewati pre-processing antara lain Resize, Grayscale, Median Filter dan
CLAHE. Selanjutnya, ekstraksi fitur akan dilakukan menggunakan metode Gray Level Cooccurrence Matrix (GLCM). Matriks dihitung dengan mempertimbangkan kombinasi
empat sudut (0, 45, 90, 135) dan jarak (1, 2, 4, 8, 16, 32, 64, 128). Nilai ekstraksi fitur
GLCM ini kemudian dijadikan sebagai masukkan metode Probabilistic Neural Network.
Hasil dari penelitian ini yaitu identifikasi Covid-19 dapat dilakukan dengan
menggunakan ekstraksi fitur GLCM dan Probabilistic Neural Network sebagai metode
klasifikasi dimana pemilihan parameter smoothing terbukti mempengaruhi akurasi
klasifikasi. Pengujian pada klasifikasi citra normal, Covid-19, Bacterial Pneumonia dan
Viral Pneumonia menghasilkan akurasi 73% pada parameter smoothing 0.9. Skenario
pengujian lain juga dilakukan dengan hasil tertinggi diperoleh pada klasifikasi citra
Normal dan Covid yang menghasilkan akurasi 100%.
Kata Kunci: Covid-19, GLCM, PNNABSTRACT
The pandemic caused by the novel coronavirus has changed the behavior of every
individual around the world. Starting from social restrictions, working from home,
wearing masks when traveling and bringing hand sanitizer. This corona virus causes a
disease with symptoms similar to cases of pneumonia. Its emergence as a new disease and
its resemblance to pneumonia have caused medical workers to be overwhelmed in dealing
with this pandemic. In general, testing is carried out using the Real-Time Reverse
Transcription-Polymerase Chain Reaction (RT-PCR) technique, but identification of
radiological images can also be considered a tool for testing Covid-19.
This study uses X-Ray Thorax imagery to identify Covid-19 infection. The image will
go through pre-processing including Resize, Grayscale, Median Filter, and CLAHE.
Furthermore, feature extraction will be carried out using the Gray Level Co-occurrence
Matrix (GLCM) method. The matrix is calculated by the combination of four angles (0, 45,
90, 135) and distances (1, 2, 4, 8, 16, 32, 64, 128). This GLCM feature extraction value is
then used as input for the Probabilistic Neural Network method.
The results of this study are the identification of Covid-19 can be done using GLCM
feature extraction and Probabilistic Neural Network as a classification method where the
selection of smoothing parameters is proven to affect classification accuracy. The normal
image classification test, Covid-19, Bacterial Pneumonia, and Viral Pneumonia resulted in
an accuracy of 73% at the smoothing parameter of 0.9. Other test scenarios were also
carried out with the highest results obtained in the Normal and Covid image classifications
which resulted in 100% accuracy.
Keywords: Covid-19, GLCM, PNN

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Covid-19, GLCM, PNN
Subjects:	Z Bibliography. Library Science. Information Resources > ZA Information resources
Divisions:	Faculty of Engineering, Science and Mathematics > School of Engineering Sciences
Depositing User:	Eko Yuli
Date Deposited:	15 Feb 2022 03:32
Last Modified:	15 Aug 2022 03:15
URI:	http://eprints.upnyk.ac.id/id/eprint/28376

Actions (login required)

View Item